cosmology with gravitational lens time delays › ... › uploads › 2017 › 02 › suyu.pdf ·...

Cosmology with Gravitational Lens Time Delays

Sherry Suyu Max Planck Institute for Astrophysics

Academia Sinica Institute of Astronomy and Astrophysics Technical University of Munich

@ Cosmology on Safari, February 13, 2017

H0, a key parameter

Tension? New physics? Need more precise & accurate H0

Hubble constant H0 - age, size of the Universe - expansion rate

Need Independent methods to overcome systematics, especially the unknown unknowns

Riess+16

[Riess et al. 2016]

3

Strong Gravitational Lens with Active Galactic Nucleus

Active galactic nucleus (AGN) in the source from accretion of material onto supermassive black hole:

Light emitted from AGN changes in time (“flickers”)

4

Gravitational Lens Time Delays

[Fassnacht et al. 1999, 2002] Movie Credits: S. H. Suyu, C. D. Fassnacht, NRAO/AUI/NSF

5

Gravitational Lens Time DelaysTime delay:

Time-delay distance:

Obtain from lens mass

model

Advantages:- simple geometry & well-tested physics - one-step physical measurement of a cosmological distance

For cosmography, need: (1) time delays (2) lens mass model(3) mass along line of sight

H0LiCOWH0 Lenses in COSMOGRAIL’s Wellspring

B1608+656 RXJ1131-1231

6[Suyu et al. 2017]

HE0435-1223 WFI2033-4723 HE1104-1805

H0 to <3.5%

precision

H0LiCOWers

H0LiCOWH0 Lenses in COSMOGRAIL’s Wellspring

B1608+656 RXJ1131-1231

8

HE0435-1223 WFI2033-4723 HE1104-1805

ongoing

completed[Suyu et al. 2010, 2013, 2014, 2017, Rusu et al. 2017, Sluse et al. 2017, Wong et al. 2017, Bonvin et al. 2017]

H0LiCOWH0 Lenses in COSMOGRAIL’s Wellspring

B1608+656 RXJ1131-1231

9

HE0435-1223 WFI2033-4723 HE1104-1805

ongoing

focus

completed[Suyu et al. 2010, 2013, 2014, 2017, Rusu et al. 2017, Sluse et al. 2017, Wong et al. 2017, Bonvin et al. 2017]

10

Time Delays

[Cosmological Monitoring of Gravitational Lenses]

EPFL: G. Meylan, F. Courbin, V. Bonvin, M. Tewes, Y. Revaz, N. Cantale, C. Faure, A. Eigenbrod, C. Vuissoz IIA Bangalore: T. Prabhu, C.S. Stalin, R. Kumar, D. Sahu Univ. Liège: D. Sluse, P. Magain, E. Eulaers, V. Chantry UzAS Tashkent: I. Asfandiyarov Univ. Zürich: P. Saha, J. Coles Univ. Nottingham: S. Dye Now also in close collaboration (monitoring, microlensing) with: C. Kochanek, A. Mosquera (Ohio), C. Morgan, C. MacLeod, L. Hainline (USNA)

• monitoring lensed quasars since 2004 in the optical • expect to have delays with a few percent error for ~10 lenses

11

Time Delays

13-year light curve of HE0435-1223Time delay with 6.5% uncertainty [Bonvin, Courbin, Suyu et al. 2017]

COSMOGRAIL monitoring

Lens environmentHE0435 has nearby mass structures at different redshifts[e.g., Morgan et al. 2005, Momcheva et al. 2015]

Wide-field spectroscopy for group identification [Sluse, Sonnenfeld, Rumbaugh et al. 2017]Wide-field imaging to get external mass distribution [Rusu, Fassnacht, Sluse et al. 2017]

Lens environmentHE0435 has nearby mass structures at different redshifts[e.g., Morgan et al. 2005, Momcheva et al. 2015]

Wide-field spectroscopy for group identification [Sluse, Sonnenfeld, Rumbaugh et al. 2017]Wide-field imaging to get external mass distribution [Rusu, Fassnacht, Sluse et al. 2017]

zL=0.4546

14

Lens mass modeling

[Wong, Suyu, Auger et al. 2017]

• New lens modeling capabilities developed in GLEE [Suyu & Halkola 2010]: • PSF reconstruction • multi-lens plane

modeling [Suyu et al. in prep.]

• Modeled strong lens including up to 5nearest neighboringgalaxies

HST imaging in 3 bands

15

Blind analysis in action

[Wong, Suyu, Auger et al. 2017]

• Blind analysis to avoid confirmation bias

• throughout analysis, cosmological distancesand parameters are offset from median, i.e., blinded

16

Blind analysis in action

• Blind analysis to avoid confirmation bias

• throughout analysis, cosmological distancesand parameters are offset from median, i.e., blinded

• collaboration agreed: when unblind, publish DΔt without modification

• scheduled unblinding telecon on June 2

• scheduled again for June 16

[Wong, Suyu, Auger et al. 2017]

17

Time-delay distance of HE0435

[Wong, Suyu, Auger et al. 2017]

DΔt = 2612 Mpc +208-191

Analytic fit for DΔt

x = D�t/(1Mpc)where

�D = 653.9µD = 7.5793�D = 0.10312

Can be combined with any other probe

18

H0 from 3 strong lenses

[Bonvin, Courbin, Suyu et al. 2017]

H0 ∈ [0,150] km/s/Mpc Ωm = 1 - ΩΛ ∈ [0,1] w = -1

H0 with 3.8% precision for flat ΛCDM

19

Cosmography with 3 strong lenses

Riess+16

3 LensesH0LiCOW+COSMOGRAIL

2 LensesH0LiCOW+COSMOGRAIL

[Riess et al. 2016]

20

Lensing REALLY helps

[Bonvin, Courbin, Suyu et al. 2017]

21

Looking forward

• 2 more H0LiCOW lenses • follow-up data mostly in-hand

and partly pending • HST WFC3 image show

Einstein ring

Expectation from H0LiCOW: <3.5% uncertainty on H0 for most cosmologies with w=-1(HST proposal aim of 3.8% achieved with 3/5 lenses thanks to the improved time delays from COSMOGRAIL)

WFI2033-4723 HE1104-1805

[Suyu, Bonvin, Courbin et al. 2017]

Towards hundreds of lenses

22

Hyper Suprime-Cam Survey

Dark Energy Survey

8m Subaru Telescope Mauna Kea, Hawaii

4m Blanco Telescope, CTIO, Chile

• 1400 deg2 with ilimit~26 • 2014-2019 • expect ~600 lenses

[Oguri & Marshall 2010]

• 5000 deg2 with ilimit~24 • 2012-2017 • expect ~1100 lenses

[Oguri & Marshall 2010]

STRong-lensing Insights into Dark Energy Survey(PI: Treu)

Kilo Degree Survey

2.6m VLT Survey Telescope, Paranal, Chile

• 1500 deg2 with rlimit~25 • 2011-~2018

23

Lensed quasars from the ground

[Anguita et al. 2009]

Space (HST) Ground (Subaru)

How to tell if this is a lens? Use configuration of blended images and lens modeling

New hunter for lensed quasars

24

Probably a lens Probably NOT a lens

James Chan has developed an automated and fast algorithm to use multiband imaging for lens classification [Chan, Suyu et al. 2015]

New hunter for lensed quasars

25

重⼒力和影像探奧號Chung-li He In-hsiang Tan Ao Hao

CHITAH

Future ProspectsCurrent/future surveys including HSC, DES, KiDS, Euclid,LSST and WFIRST will provide ~10,000 lensed quasars and ~100 lensed supernovae [Oguri & Marshall 2010]

[Jee, Komatsu, Suyu, Huterer 2016]

27

Summary• Time-delay distances DΔt of each lens can be measured with

uncertainties of ~5-9% including systematics • Recent blind analysis of HE0435-1223 using time delays from

COSMOGRAIL, wide-field imaging/spectroscopy, HST imaging, and lens velocity dispersion from Keck

• With 3 time-delay lenses: H0=71.9 km/s/Mpc in flat ΛCDM • H0LiCOW: H0 to <3.5% precision from 5 lenses • Search is underway to find new lenses in HSC and DES

surveys with new lens hunter CHITAH and other methods • Current and future surveys will have thousands of new time-

delay lenses, providing an independent and competitive probe of cosmology

+2.4-3.0

Thank you!

28